Brief background

Alsulaiman et al. (2021) suggest that Phox2a is preferentially expressed in lamina 1 projection neurons of the spinal cord. These projection neurons have synapses on the ipsilateral side but send signals to the brain on the contralateral side In the mouse and human single cell/nuclear RNA seq datasets, we have neurons from the dorsal horn of the spinal cord that could be projection of interneurons. Therefore, our research questions are could we use this Phox2a gene to identify lamina 1 projection neurons within the clusters, how much Phox2a is expressed in the mouse SDH vs DDH, are there more excitatory or inhibitory projection neurons, are there any genes that are co-expressed with Phox2a - can we use it to estimate which neurons in general are located in lamina 1, what is the distribution of the Phox2a gene across clusters and cells, how many cells in the SDH and DDH highly express Phox2a, can we re-cluster the cells based on Phox2a expression.

Based on the information and research questions I hypothesize that:
- Phox2a expression patterns will have a range/variance that suggests the presence of a different subtype of dorsal horn neurons
- There will be higher average Phox2a expression in the mouse SDH than the mouse DDH.
- It is possible that some clusters may have more Phox2a expression than others because they were formed by looking for differentially expressed genes.
- Phox2a is more highly expressed in mouse and human dorsal horn cells labeled as excitatory rather than those labeled as inhibitory.
- We will need to look more into the protein product of Phox2a and its roles know if it can be used to preferentially identify other lamina 1 neurons.
- Phox2a more higly expressed than NMDAR subunit genes

Results

We will use the clusterpooling pipeline on 4 different analysis projects:
- mouse SDH vs DDH excitatory vs inhibitiory
- mouse SDH vs DDH
- human excitatory vs inhibitiory
- human all pooled together

Mouse SDH vs DDH excitatory vs inhibitiory

We will begin with the mouse data and looking at SDH vs DDH excitatory vs inhibitory neurons. The full dot plot and pooled dot plots are bellow.

The full dot plot Fig 1. NMDAR subunit and Phox2a RNA average expression and percent expressed in SDH and DDH neurononal clusters from the mouse dataset

The Pooled Dotplot Fig 2. The pooled dot plot showing the average expression and percent expressed of the NMDAR subunits genes and Phox2a gene levels for SDH excitatory and inhibitory neurons and DDH excitatory and inhibitory neurons.

Cluster range plot Fig 3. The raw log scaled transcript levels of individual cells in each color coded cluster for the NMDAR subunit genes and the Phox2a gene for each of the 4 groups.

Based on the dot plot and the cluster range plot it seems that not a lot of cells express the Phox2a gene. Could it be more localized in the cytoplasm. In fact it seems that the DDH excitatory neurons have higher percent expressed. Three clusters by eye have higher percent expressed than any of the other clusters. Those clusters are Excit 21, 23 and 29. According to Russ et al. (2021) the highly expressed genes in these clusters are: Lmx1b, Zfhx3, Nms, Lypd1 for Excit 21; Lmx1b, Nfib, Cep112, Cdh23, and Satb1 for Excit 23; and Onecut2 and Pmfbp1 for Excit 29. We will have to do some extra research to see if these genes relate to Phox2a.

If we look at Seqseek and look for other Phox genes as well what patterns emerge. Neither Phox2a nor Phox2b are expressed highly in a large amount of cells.

SeqSeek Phox2a expression

SeqSeek Phox2b expression

Mouse SDH vs DDH

Based on the above results we can skip this step, and consider if Phox2a is present in human lamina 1 projection neurons and not mice lamina 1 projection neurons.

Human excitatory vs inhibitiory

A quick visit to the shiny app from Yadav et al. (2022) shows that there are very little transcript levels of Phox2a/b in the any of the cells.

Human single nuclear RNA seq shiny app Grin1, Phox2a, Phox2b expression

Human all pooled together

We can skip this step too.

Discussion/conclusion

Unfortuantley, non of the hypotheses are true/valid. Both the mouse and human data sets show that Phox2a/Phox2b transcript levels is very low across the majority of neuronal clusters in the SDH or DDH. Very low transcript levels means that it would be hard for us to draw conclusions because a lot of factors can contribute to cells having low transcript levels. Could it be that Phox2a is only expressed when a projection signal is being sent, or is it not localized in the nucleus? Low transcript levels could be a result of strong regulation or the technique itself. For instance, what was the Cre or BAC protocol doing that distinctly differes from the single cell RNA seq experiments. Albeit we are comparing two methods and there could be flaws with both. There are a few mouse DDH clusters that have a decent percent expressed of Phox2a, which could be investigated further in the future. The only conclusion that can be drawn is that it is unlikely that Phox2a can be used with these single cell/nuclear RNA seq mouse and human datasets to identify projection neurons.